Tito and I were just talking a few minutes ago about the $10
thermocycler. I know that Joseph Jackson is working on commercializing
it, but the DIY crowd can bypass the patents and just go make it. What
we need at the moment is for a few people to read through the paper,
make some notes on the design, post the notes on the design, and then
come up with some arbitrary schedule for the project in terms of
milestones. I wouldn't mind contributing to the design.
$10 thermocycler paper-
I'll be sure to post some notes of my own in a while, but I will
readily admit that my electronics skill has been rusty, now that I can
hardly sneak a soldering iron into my dorm. ;-)
You would only want to do microfluidic-PCR if your specific project
required very small volumes. For general purposes I would stick to
"regular PCR" (20-50uL). The microfluidic platform just adds an
unnecessary level of complexity to something that is fairly simple.
But if you're actually interested in microfluidics, then by all means go for it!
> A while back there was an article about making general
> microfluidics devices using Shrinky-Dinks
> (http://www.labplusinternational.com/index.php?id=2155) that sounds very
When that paper first came out I tried it out in my lab. I didn't get
very good results. The main problem was that the shrinky material
didn't shrink uniformly. I tried a number of different ovens around
campus, including some convection ovens (recommended by the authors),
but the channels always ended up warped. I'm not sure if the problem
was due to the material or just uneven heating. Other people may have
improved on the technique since the original publication. If you're
interested in the shrinky-dink-microfluidics you may want to browse
the journal "Lab on a Chip" for more recent protocols.
This list is great btw. Every day a couple of things turn up that make me go "holy crap!"
http://heybryan.org/~bbishop/docs/diybio.png (warning: obscenities)
That about sums it up, Nick.
link is down
> You received this message because you are subscribed to the Google Groups "DIYbio" group.
> To post to this group, send email to diy...@googlegroups.com
> To unsubscribe from this group, send email to diybio+un...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/diybio?hl=en
Rochester Institute of Technology
College of Science, Biotechnology/Bioinformatics
The thermal mass of TE coolers needs considering. Running them dumps heat right into them, so switching
involves delay while the heat flows. The hot side heat needs to all go some where before
that spot can be cool. I like the idea of running the coolers steady state and
flipping open different doors to allow air flow past them or past heaters or ambient
to rapidly change heat flow to or from samples in heat finned vials surrounded by
flowing air. Small door flippers can be made from plastic piezo electric film actuator material
that is inexpensive and fairly low power consumption, (think battery operation, solar cell operation, or just
low waste heat generation in your room)
I don't see how a price of approx. $20 will be any more helpful for DIYbio
budgets than one that is $80 with time saving conveniences like thermocycling
12 samples at once, not having to super clean the special shaped tube
in place in the machine or re-thread a new one in after running a PCR on just one sample.
The TX A&M design with triangular tube path does have very low power consumption
good for battery operation, but I can easily see a machine based on air heat transfer
to vials having low enough power consumption to run on a small solar cell for
jungle/remote field applications. Plus the TX A&M design is a patented process, so
not available to produce even in kit form, where classic PCR can be DIY low cost/easy
and manufactured as an open hardware kit you don't have to hack bits of metal to get.
I would say calibration is a big disadvantage.... what happens if a
guy down the hall/street in a different lab wants to borrow your PCR
machine, and he happens to have a window or Air Conditioner raising or
lowering his lab's temp. I wouldn't want to have to readjust screws
and stuff like that all the time. What about if I'm doing a PCR run,
and the central air conditioning turns on...
Unless ambient temp had a linear effect on all three heating
blocks/areas, this design suffers significantly in that area.
> TL;DR The $10 pocket PCR design could be adapted by us if we put our
> heads together. The $10 dollar claim is a little extreme but it's
> easily sub $30 factoring in shipping if you have a soldering iron,
> wire, a way to machine aluminum and a thermocouple already. If
> someone has a way to machine the blocks, I'd gladly trade a soldered
> circuit, or maybe we can put our heads together and think of
> premanufactured parts that could serve the role of the aluminum
> -Matt Conway
> B.S. Chemical Engineering '12
> University of Maryland
> On Sep 12, 2:07 pm, John Griessen <j...@industromatic.com> wrote:
>> On 09/12/2011 10:10 AM, Dakota Hamill wrote:
>> > although you need a way to move the heat off of one side to maintain the temperature gradient.
>> The thermal mass of TE coolers needs considering. Running them dumps heat right into them, so switching
>> involves delay while the heat flows. The hot side heat needs to all go some where before
>> that spot can be cool. I like the idea of running the coolers steady state and
>> flipping open different doors to allow air flow past them or past heaters or ambient
>> to rapidly change heat flow to or from samples in heat finned vials surrounded by
>> flowing air. Small door flippers can be made from plastic piezo electric film actuator material
>> that is inexpensive and fairly low power consumption, (think battery operation, solar cell operation, or just
>> low waste heat generation in your room)
> You received this message because you are subscribed to the Google Groups "DIYbio" group.
> To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to diybio+un...@googlegroups.com.
> For more options, visit this group at http://groups.google.com/group/diybio?hl=en.
I'm no expert, but I tried to amplify a piece of mRNA and didn't get
what I wanted. I'm going to try again by adjusting the annealing temp
a few degrees colder, then I'll try adjusting the Mg2+ concentration.
A few degrees can make the difference between specific and
non-specific binding. In the above case, I added linker regions to the
ends of my primers, so they're not completely specific. Decreasing the
temp a few degrees could prevent the non-specific regions from
dissociating the whole primer, and getting me the sequence that I want
to avoid getting synthesized.
How about using CAD such as HeeksCAD to design some parts that can be cut out inexpensively?
Pre-made parts for something else will always be crude hacks stuck together with JB-Weld and
not very pretty, not anything that can be made inexpensively as a full assembly, just a
penny pinching hobbyist kludge.
How about using our heads to come up with a moderately priced path to a < $100
thermal cycler system made mostly of flat parts so they can be sent
around to different developers and testers by low cost first class mail?
I'm thinking of mechanical parts that fit together with press-on fit, screws,
silicone caulk. No sawing, drilling or measuring to assemble a kit.
I use HeeksCAD on linux, the main developers run it on windows, and it can probably
run on OS-X with a little effort.
I can spend small slivers of time to help others get up to speed for making things. Who's interested?
In an ideal PCR, temperature can be pretty flakey. Sadly, all too often you need to optimise your sharpen bands or to get bands at all, and a well defined temperature profile can be critical.
The main issue addressed is non-specific binding, including primer dimers. However, primer hairpins and complex DNA (i.e. Secondary structures) can benefit from tightly controlled annealing conditions too.
So, depends: if you're doing off-the-shelf reactions designed for bush conditions, maybe not. Research or Development conditions, probably you do need good accuracy.
231.313.9062 // @100ideas // sent from my rotary phone
>> On Sep 14, 12:23 pm, Jonathan Cline<jncl...@gmail.com> wrote:
>> In fact what comes to mind is if water quality might be more
>> important for the DIYer than the temperature accuracy of the
I've heard before that the purity requirement for soda pop was more stringent than any natural water.
Wouldn't it be funny if Dasani water (by Coca Cola) and some added drops of Gatorade was just right
for electrophoresis resistance adjusting.
What a disgustingly wasteful way to make drinking water! Ow. Now I know to avoid that if I'm ever unlucky enough to need bottled water over yonder.
When it comes to ions and DNA, Ca, Mg and Mn spring to mind as most relevant catalytic ions, but that's top of the head stuff. Others will of course play roles with specific enzymes, like cobalt.
Adding cheating agents such as citrate or edta will sequester compatible ions one-for-one, so dose by molarity not weight.
I'm interested in seeing sketches of how a hot-air based cycler would
look. What I really mean is how the intake and exhaust manifolds would
look, because I think they should promote even flow (at least on a
short term average) amongst all the sample tubes.
Maybe we could directly use the OpenPCR firmware, modified to switch a
heater instead of a peltier.
Here's a rough sketch of a temp controlled chamber for a carousel holding
vials to spin in.
The turbulence of spinning at least 2 revs per second (120RPM)
along with minimal solid contact with vials will keep all vials the same
as far as heat exchange.
Here's a diagram of an easy to fabricate tube well shape that minimizes
contact with the vials:
Controlling the doors with continuous position might not be
possible, but snapping them open and shut often seems easy.
Steady state would not require doors to move much, just when
changing/ramping temperatures -- so they would not wear out.
The material I am thinking of for doors is a flexible
plastic composite with piezo effects. I've not tried this yet,
and not sure of stiffness of the material, so the doors that
stick out in the air flow might not work that way if the material
is too flimsy. Some design around that will work though --
that piezo material can be used to make a fan.
spinning the tubes to equalize temps makes sense
> The turbulence of spinning at least 2 revs per second (120RPM)
> along with minimal solid contact with vials will keep all vials the same
> as far as heat exchange.
> Here's a diagram of an easy to fabricate tube well shape that minimizes
> contact with the vials:
pretty rough sketch, is the center the tube holder, or the 6 holes
around the center for the tubes?
> Controlling the doors with continuous position might not be
> possible, but snapping them open and shut often seems easy.
> Steady state would not require doors to move much, just when
> changing/ramping temperatures -- so they would not wear out.
> The material I am thinking of for doors is a flexible
> plastic composite with piezo effects. I've not tried this yet,
> and not sure of stiffness of the material, so the doors that
> stick out in the air flow might not work that way if the material
> is too flimsy. Some design around that will work though --
> that piezo material can be used to make a fan.
doors for what? I was imagining PWMing the heater to change temps,
since room temp is way out of the range for cycling. Unless you mean
to add room temp air to get the temp close to the set-point, then turn
the heater on to maintain the set point (can't just cool it down then
heat it back up, needs to hold at a temp for some time)
Also if there were multiple fans, one for fresh air, one for heated, I
don't see the need for a door, just that one of the fans would always
be on, maintaining positive pressure.
> pretty rough sketch, is the center the tube holder, or the 6 holes
> around the center for the tubes?
center is tube carousel holder
> doors for what?
Doors change flow through hot zone or to ambient==cooler.
I was imagining PWMing the heater to change temps,
Sure that too, but take away the slowness of thermal masses changing temps
and just throw doors to get different temp and ramp the other way.
> Also if there were multiple fans,
Carousel is the fan. Simplicity.
hmm, really? OK, I'll look for some turbine 3D design files online,
maybe thingiverse has something.
It does not need a turbine shape. It will be a functional fan with or without
vials set in place. If some vials missing, the holes will have plenty of air drag
to pull air along as it spins.
This design is something that can be mostly laser cut from sheet plastic
for easy manufacturing and low costs. The housing can be made on a rep-rap
or makerbot. This sketch is about holding cylindrical vials that have caps
that are wider with no ability to deal with plates, but future
possibility to work as an incubator
and liquid handler and transmission densitometer with many of the same parts.
how would you direct which way the air will flow, without turbine
features? I.e. what if the air gets pushed from in the cycler past the
heater, to the outside (heating the room, rather than the tubes)
also, 2 RPMs seems too slow, but I wouldn't want to depend on faster
speeds because it could cause chemical or protein gradients in the
tube (mayyyybbeeeee, I have no idea how fast it would have to spin for
this to happen)
> This design is something that can be mostly laser cut from sheet plastic
> for easy manufacturing and low costs. The housing can be made on a rep-rap
> or makerbot. This sketch is about holding cylindrical vials that have caps
> that are wider with no ability to deal with plates, but future
> possibility to work as an incubator
> and liquid handler and transmission densitometer with many of the same
The doors control that. Sure, it's just a theory at the moment.
I'll be testing it out though...
> also, 2 RPMs seems too slow, b
2 revs per second is a good low G-force speed.
Like an impeller pump, and by opening and closing ports (doors):
> The straight vanes look innefficient, is that so?
We want low flow, low noise, and don't need it piped into a tube.
> This looks better:
Having the carousel for vials be the impeller is more efficient.
When your goal is to exchange heat with the vials on the carousel,
and you do not care about directing flow in a tube. All the flow we
need is turbulent mostly to even out temperatures, and enough to
exchange heat, which is easy and doesn't need any optimizing or
calculations even. Pumps can be optimized, but we don't need a pump,
we just need heat flow.
I don't follow 'not needing a pump/fan', how will air surrounding a
hot coil be directed past tubes needing heat? There needs to be a fan,
in my opinion the vanes should look like a helical gear, to draw air
past the upward and past the tubes. If the carousel is only to even
out the temps, then a fan behind the coil would be needed to push any
air through the system.
sure for things like lightbulb PCR, but my understanding of what John
and I have been talking about is akin to pointing a hair blow dryer
towards and then away from the PCR tubes.
> enough to get the job done. i'm sure that both would work, but adding a
> bunch of equipment not stricly needed might not be the way to go. i guess it
> depends on what goal you're trying to achieve.
an extra 100mm computer fan isn't 'a bunch' of stuff though, its about
$4 new (if that)
> housing it inside a computer case, like the OpenPCR, sure...
openPCR isn't housed in a computer case, its in laser-cut luan.
> I think that what you're missing is that PCRs often rely upon reactor
> geometry to achieve what you think needs to be done by a fan or pump.
hmm, that seems to rely more heavily on thermal physics and
engineering... not my idea of a first iteration at a new type of PCR
> Here, this will help you understand (and she's got a damned sexy voice
> anyway so worth listening to).
> Rapid PCR Thermocycling using Microscale Thermal Convection
> It's probably the best instructional video on the subject matter that I've
Thanks, I'll check it out
by opening or shutting doors at which there is a pressure difference
from fan action and flow will happen as soon as doors open.
There needs to be a fan,
There is a fan. The carousel *is* a fan impeller in a housing as I've sketched
in prev. emails.
> in my opinion the vanes should look like a helical gear, to draw air
> past the upward and past the tubes.
OK, but with a separate turbine is a different design without
much reuse of parts and I don't see why it's necessary -- or how it could
be made for a profit as open hardware.
The sketches tell a lot. Can't answer much more for a couple of days.
Busy with a concrete pour Tues.
I guess I'm just thinking that air wouldn't circulate through the
heater duct without the vanes/blades of the fan (which happens to be
the carousel) being directional. The non-helical fan I mentioned from
thingiverse looks like a boat paddlewheel, doesn't seem non-directed
air flow is the best way to get high ramp rates.
> There needs to be a fan,
> There is a fan. The carousel *is* a fan impeller in a housing as I've
> in prev. emails.
How fast would it have to spin to generate sufficient air mass and
heat transfer though? I don't think 2 RPMs are going to cut it.
>> in my opinion the vanes should look like a helical gear, to draw air
>> past the upward and past the tubes.
> OK, but with a separate turbine is a different design without
> much reuse of parts and I don't see why it's necessary -- or how it could
> be made for a profit as open hardware.
one fan versus two is not a profit killer, I'm not sure how you see
that, but I wasn't saying its necessary, but if you want to spin at 2
RPM to even out temp eddies, then yeah I think another fan to actually
blow air around is needed.
> The sketches tell a lot. Can't answer much more for a couple of days.
> Busy with a concrete pour Tues.
Then the next steps are
1. make prototype stuff
2. test prototype stuff
3. repeat til working as few times as possible based on hunches and experience and inductive logic.
No heat flow calcs. Takes too long. The container is one of the worst heat flow
gaps being plastic. Heater behind a door is a no brainer for super rapid air temp change
copied from GM cars. There will be no problem of transferring heat at all. The hot door
will have to be shut 2 seconds after the start of a ramp up is my bet. And during the ramp it
will open and shut every 3/4 second likely. That will all be part of a heuristic
semi-PID control with a little nonlinear/nonintegral preboost and it will tune up easily.
Next comes building parts, testing parts.
"CyclerCan" was air-based. Got air gun plus cooler fan.
The heater and chamber thermal mass and convection air is such a loosely connected system
I doubt a simulation would be helpful as first step. Especially when you know you can just dump some
heat energy and most of it will stay in the machine for a couple of minutes...then just
control away by varying door openings while hot zone is open and cooling off, or shut and
steady state with little heat input.
have to have 20 iterations of the
> same hardware to get results.
20? No, 3 iterations of 1/2 of the design are normal for design of anything good.
And half is good the first try usually.
If you're building equipment like this and PDE is too much
> for you, you really might consider spending more time learning.
I'm just fine at partials and time domain calcs, thanks for sharing, but will
you be volunteering any effort on this?
If you keep the thermal mass of the heater very low that could work. I really like the doors for access to ambient air
so I don't need a TE cooler. Once you control some doors for ambient air, then using some for heat is a natural also.
I always have a design goal to perform to a quantitative level rather than "pretty fast". I'd like to offer programmable ramp
times of 1 deg C per second. Maybe the heater doors could be left off, but the ambient doors are going to be needed
to have any benefit of air heat exchange.
I don't think
> anyone has mentiones it yet but Corbett solved this particular problem a while ago with the RotorGene qPCR. I don't think it has
> any other moving parts other than the rotor/fan that carries the PCR tubes.
Dang! They have glitzy industrial designed packages around their equip and really brag about a carousel used to hold vials. That
means they may have patented it...hmmm...
number 6814934 and . 6,787,338; 7,238,321; 7,081,226; 6,174,670; 6,245,514; 6,569,627; 6,303,305;
6,503,720; 5,871,908; 6,691,041; 7,387,887; 7,273,749; 7,160,998; U.S. Patent Application Nos. 2003/0224434, and 2006-0019253, and
PCT Patent Application No. WO 2007/035806, and
all continuations and divisionals, and corresponding claims in patents and patent applications outside the United States, owned by
the University of Utah Research Foundation, Idaho Technology, Inc., Evotec
Biosystems GmbH, and/or Roche Diagnostics GmbH.
That's a lot to read. Will take a while...
Didn't read any of it yet.
Do you think luan (the thin plywood) would be OK for a prototype?
> You received this message because you are subscribed to the Google Groups
> "DIYbio" group.
> To post to this group, send email to diy...@googlegroups.com.
> To unsubscribe from this group, send email to
> For more options, visit this group at
Sure, why not? I bet you're thinking of laser cuttables, right?
Get your smoke exhaust vent going and tell us about it!
You can make 3D shapes by stacking layers that are glued together
Try using HeeksCAD for the 3D and I will be able to help with ideas
and versions. Plus anyone else since it is free OSS. And it's good!
It has parameterized 2D sketches and python scripting.
Yep thinkin laser cut pieces... Ideas for easy to source heaters... Would like to avoid using a power FET for now, if possible.
Will also need to think about what to use to actuate the doors, cheap and quick linear actuator, but from where/what?
Take a look at this ubuntu linux/Mac driver for epilog lasers:
Ideas for easy to source heaters... Would like to avoid using a power FET